Here i got parsed literature site https://avidreaders.ru for about 10.000.000 sentences from russian books and make sentence vector embeddings from them using Mistral open API.

Embeddings got resized from 1024 to 256 dimensions using python scikit-learn PCA method.

Word embeddings are a way of representing words as vectors in a multi-dimensional space, where the distance and direction between vectors reflect the similarity and relationships among the corresponding words.